Short answer: No. You should always be wary of trying to apply evolutionary models to achieve specific real-world goals. Given the past history here, I'll go out on a limb and say that it's always a mistake.

Even if it was the right choice, these sorts of things have a dizzying array of unintuitive results and even the academics involved are still working out a lot of the implications. In particular, we don't have a good understanding of how applicable these results (and the wider predictions of HBE) are to non-foraging human societies.

Fairy Nuff.

I think we do know of plenty of anti-patterns (ie concrete monstrosities, lack of local walkable amenities) so urban planning is still a Good Thing. If done right.

It seems to me that changing the environment to manipulate human behaviour ultimately creates stress (in a very literal sense), which is likely to undermine the goal.

Except that environmental change to manipulate human behavior is ubiquitous. Consider traffic lights, fences, handrails, etc.

The only question is what behavior we’re trying to achieve and how we’re manipulating the environment to achieve it.

Not necessarily - the change might be to make an environment that's less stressful to humans (e.g. humans might like an environment where they can see what's coming and avoid being ambushed).